Can AI-Generated Art Be Protected? Exploring Copyright and Privacy Risks

As AIGC technology advances, more people use AI to create images and videos, even selling them on major platforms. A recent AI‑generated NFT titled “The First 5000 Days” sold for over $600,000, prompting a copyright dispute as Larva Labs claims the work incorporates protected CryptoPunks images without permission.

Copyright protects original works; only original creations are eligible. AI‑generated works can infringe copyright when the training data includes protected content, creating a high risk of infringement.

Creators may feel outraged when their work is misused, while users of AI‑generated content need clarity on the legality of their actions.

Although some argue that massive training datasets and complex algorithms reduce infringement risk, studies on models like Stable Diffusion show a 1.88% chance of generating content with over 50% similarity to original works, indicating a non‑negligible risk.

The EU has introduced new copyright regulations that address AI‑created works, defining ownership and usage rules. In China, the Copyright Law allows AI‑generated works that meet the criteria of originality to be protected, though ownership may still be contested.

Creators should use legally sourced materials, build their own datasets, and avoid over‑reliance on AI, ensuring the final output reflects their own creativity.

Beyond copyright, AI raises privacy concerns: models may use uploaded personal photos without consent, leading to potential data leaks and misuse such as deep‑fakes for fraud or misinformation.

In summary, AI technology itself is neutral; its impact depends on how it is applied. Users must recognize and mitigate risks like copyright infringement and privacy violations, define clear usage boundaries, and strive for responsible AI adoption.